Symantec’s first “State of the Datacentre” report, published last week, highlights the ICT pressure-points facing datacentres.
Datacentre managers are constantly finding themselves being asked to do more with less, but technology solutions are only going part of the way towards easing this burden. At the same time, a shortage of qualified staff remains a continuing concern.
As businesses become increasingly reliant on ICT — for both information and transaction-processing — more applications are being upgraded to “mission-critical” status.
This, in turn, has led to businesses becoming more demanding when it comes to service-level agreements (SLAs).
Datacentres are also growing and becoming more complex, often including a diverse collection of computers, but their budgets are not keeping pace with this growth, says the Symantec report.
The report quotes statistics for the Asia-Pacific and Japan (APJ) region which were extracted from a worldwide survey and highlight the mismatch between demand and capability in this part of the world.
Of APJ datacentre managers polled, “38% reveal they have more difficulty meeting SLAs over the past two years,” according to a presentation by Paul Lancaster, systems engineering manager for Australia and New Zealand. A whopping 85% stated that SLA expectations had increased over the last two years, and 28% reported SLA expectations had increased rapidly.
Most IT managers in the region saw expenditure growth of between 5% and 9% per annum, but with budget growth as low as 3% a year. As a result, “a lot of IT departments are looking at consolidating the number of vendors they deal with” and adopting “standard” sets of hardware and systems software so as to make work easier to manage.
Outsourcing of the more standard parts of the datacentre operation such as back-up and storage management is also becoming more popular. The idea of “storage as a service” is a growing concept. In theory, this should free-up internal staff to concentrate on matters specific to the business and to their being upskilled as a result.
Server consolidation and virtualisation are also seen as ways to reduce complexity, although use of the latter in the AJP region lags behind the worldwide trend here. The difference is seen on the report pie-chart, with 39% worldwide “beginning to implement” virtualisation, while only 29% are doing so in the APJ region. But, while 44% worldwide are currently “discussing” virtualisation, 54% of those in the APJ region are doing so.
In Australia and New Zealand, use of virtualisation is largely confined to testing environments, rather than production, says Lancaster.
“Initially, it was seen as a big cost-cutting measure”, but then difficulties arose concerning “high availability” – keeping a virtualised application running at all times – and in the provision of disaster recovery for virtual servers.
Staff capability and staff retention are particular pressure points in the local region, says Lancaster. A good many ICT people seem to prefer to train in a relatively narrow field. “[But] we need to invest in cross-training staff… [so they can] handle several sectors of the overall workload.”
Lancaster also sees a particular need for “a management solution that gives a complete view of the datacentre and manages all hardware, and extends to all platforms.”
This exists at a high level, with suites such as CA Unicenter and Tivoli, he says. But a more consistent view of each component is needed — virtualisation, storage and so forth — at the level below that top view.